The return of ‘land’ in macro economic discourse. Wonkish
Summary. One of the most influential critics of the ideas of Piketty is Matthew Rognlie – who, to be able to write down his criticisms and following the national accounts, reintroduced the idea of ‘land’, or unproduced inputs like land and natural resources including land underlying buildings, in a neoclassical world. Herewith he undid the work of John Bates Clark, who purged ‘land’ from the concept of capital of classical economics, therewith enabling the rise of neoclassical economics. But this is not the only example of the return of land into economic discourse – land has made quite a return. Some examples:
Classical economists made a distinction between the factors of production ‘land, labour and capital’ while neoclassical economists only distinguish ‘capital and labour’ – which, as this disables any analysis of incomes related to the ownership of ‘land’ (not just land but also other unproduced inputs like oil or clean air), enabled neoclassical economists to restrict their attention to the asset side of the national balance sheet and to describe the distribution of income as a process unrelated to ‘property’ and ‘ownership’. According to economists like Mason Gaffney and Fred Harrison, purging the concept of ‘land’ from neoclassical economic discourse was a deliberate act, sponsored by people owning ‘land’, to enable exactly such a class free, a-historical, harmonious description of modern economies.. ‘Land’ has, however, made a glorious comeback. Below, I’ll point out some landmark studies.
- We can make a distinction between factors of production and factors of distribution. The asset side of the balance sheet shows the value of produced and unproduced non-labour inputs as well of the value of financial assets owned by a household, a company, a government or even an entire country, including non-tangible ones like ‘Goodwill’ and the value of patents. Though aggregating the value of these pretty heterogeneous assets is tricky side is somewhat ‘objective’. The non-financial assets are factors of production. The liability side of the balance side shows who ‘owns’ these factor. ‘Owns’ between parentheses as ownership should be understood as a bundle of rights and here for instance including the ownership of rights to interest payments of a lender who has lent money to a household, a company or the government. An example: when I do not pay my mortgage (and even when I do…) Dutch law and the mortgage contract give my bank the right to sell my house – a right which banks however sparingly use in my country. As will be clear (I hope) there are quite some historical, international and cross-sectional differences between households and companies when it comes to these ownership rights. Clearly, differences and changes in the structure and nature of ownership (including what can be owned!) influence the distribution of income between households, companies and the government. Much more elaborate on such issues: Jamie Morgan.
- Land is defined, here, as ‘unproduced assets’, which have economic value as humans use them. The obvious but, in a quantitative sense, not the most important kind of land is agricultural land. Mind: improvements (like drainage systems) are not a part of ‘land’ but of gross fixed capital. Here is a map of the ‘surface area required to power the world’ using solar cells. The point: soon as the infrastructure to tap and transport solar electricity is installed, ‘surface area’ in the neighbourhood of sun farms will rise in value, as it will be convenient to use the existent infrastructure to expand production: typically, the value of ‘land’ is not only dependent upon improvements by the owner (i.e. placing solar cells) but also upon improvements and investments by other people or by the government. To put in a ‘Piketty’, the movie ‘Once upon a time in the west’ (voiceover: ‘a manhunt, a land grab’) is about exactly this – and about the sometimes brutal nature of the distribution of ownership rights (aside – linking to this movie is not good for blogging productivity). This already holds for agricultural land but a fortiori for land in cities – which is often weirdly expensive as it is weirdly valuable, because of all the roads and houses and companies around it. Look here for $1.000.000,– parking spaces in New York. But ‘Land’ as used here also consists of oil, copper ore and, as far as I’m concerned, of clean air. Which has a cost price. Think of the costs of the catalytic converters we use in cars to prevent pollution.
- At this moment we see a revival of ‘land’ in economic theory. The magnitude of this survival is surprising and its exact timing too. But it was bound to happen. Way back in the fifties ‘land’ had become a minor asset class compared with its overriding importance only fifty years before. But as cities increased, economies grew, as out use of primary products like copper and oil grew and as ‘helicopter money’ printed by the banks fuelled house price increases but also as the average number of persons per house decreased its importance increased – and not with a little bit. Look here for some data on this process.
- Economic statisticians – often the first economists to catch up on changes in the economy! – were ‘ahead of the curve’ to acknowledge this and developed and measured (as part of those much maligned but actually pretty glorious national accounts) sectoral as well as national balance sheets, including ‘land underlying houses, other buildings and roads’ and natural resources. Which showed unequivocally that ‘land’ had become a mayor asset class (not to say: the mayor asset class) once again. In the USA. In Germany. In Sweden. In France. In the Netherlands. In the UK. It turns out that this development is not bound to national boundaries.
- Comes in: Piketty. These national accounts data are the very data which are, 1:1, the starting point for the economic history work Piketty and his comrades, who extended the official series (which often start somewhere in the seventies of the twentieth century) backwards to 1800 and even beyond. Which clearly shows the decline and rise of ‘land’ as an asset. But with, to the horror of neoclassical economists, also showed that ‘capital’, as defined on the liability side of the balance sheet, is a historical and in a sense ‘revolutionary’ construct as best exemplified by the disappearance of slaves from the sectoral and national balance sheets of the USA. From Cromwell to Abraham Lincoln and Castro revolutions and revolutionary leaders have, for better or worse, influenced and defined what we call ‘capital’. Which, of course, also means that the distribution of the spoils is influenced by these changing definitions.
- As is well-known it is pretty difficult to make a consistent estimate of the aggregated value of heterogeneous assets which are produced in different years of even periods and which are sometimes traded on the market (second hand planes) and have, hence, market prices but which sometimes ‘only’ have a (changing) production cost price and a replacement price as they are not traded by strong consensus (coastal levees). I’ll not delve into this in this post. But it is ironically to note that the favorite valuation method of economists (discounting forecasted future streams of incomes using a forecasted interest rate) is the one method not used by economic statisticians as it does not yield stable results.
- The work of Piketty c.s. was commented upon by scores of other economists. Here it comes: the arguably most influential of these, Matthew Rognlie, undid what John Bates Clark and (following him) Solow explicitly did: purging land from the neoclassical production function – but Solow can be excused a little as, in the fifties, land was indeed less important than it is nowadays (Solow, R. M. (1956), ‘A contribution to the theory of economic growth’, The quarterly journal of economics 70-1 pp. 65-94.). Impressed by the national accounts data which showed the come back of land as an asset and by the obvious economic difference between unproduced and produced capital Rognlie explicitly reintroduces ‘land’ in the neoclassical
productiondistribution function. It would indeed by ironic when, as is not impossible, Rognlie would receive the John Bates Clark medal for undoing the work Clark is most notorious for! interestingly, Clark also was the auctor intellectualis of the concept of the ‘representative consumer’, the idea that a market based society can be described as a harmonious person, without regard to class, age, wealth, power or class. Clearly, making a distinction between ‘land’ and ‘gross fixed capital’ almost necessarily introduces an ownership category like ‘house owners’ into the equation and therewith by necessity also people who do not own a house, wich demolishes the idea of a single representative consumer. Aside – many people praise him for showing the importance of ‘land’ but even a quick scan of the national accounts data (or a fast look at the Piketty graphs) shows this.
- But Piketty and Rognlie are not the only ones to focus on land. Jorda, Schularick and Taylor showed that no less than two-thirds of non-interbank lending of banks serves to finance mortgages and, hence, to quite an extent the purchase of ‘land underlying houses’. Caveat: this is net lending. The share of housing in gross lending (including the temporary financing of business purchases and international trade) is a lot smaller. Even then, the share of non-depreciating land on the balance sheets of banks increased, and increased. Land is of prime importance to the balance sheet of banks which of course means that banks have become increasingly vulnerable to land price bubbles and busts.
- And we can also point to increased coverage of house prices (including the price of new dwellings) by institutes like Eurostat and the Bank of International Settlements. Not exactly land prices – but related to it.
So, ‘land’ (as stated: including limited natural resources) is back, in the national accounts, in the discourse on inequality and distribution and in financial macro economics. Policies have to follow. One possibility: introduce a land tax, the ‘land’part of the purchasing prices of existing and new dwellings should be financed by national pension funds, the ‘dwelling part’ might be financed by banks.